Non- horizontal position of the weighing instrument: Accurate balances are fitted with a spirit level (this is even mandatory on legally verified weighing instruments) for always being able to find back the reference position after moving the instrument: this is the position in which the instrument was correctly adjusted. In the reference position the direction in which the force (the weight) is measured is the same as the direction of gravity. If, for instance, the weighing instrument is moved to a table that is not level, but is tilted 5mm on a length of 1m, and it is not set level again, the direction of measurement deviates 0.2865° from the reference position. The measured weight value W will then be cos0.2865 smaller (W1 = cos0.2865.W = 0.9999875.W), that means on a weight of 200g a 2.5mg lower readout. The zero setting also changes: the zero point is also a weight value, it is the weight of the empty weighing pan.

When in the reference position the direction of measurement exactly coincides with the direction of gravity, the result will always decrease when the instrument is tilted, when this is not the case, the error can be plus or minus. On weighing instruments with a small number of measuring steps (up to about 10000 digits) the influence of  a not horizontal position under normal circumstances is negligible, therefore these often don’t have a spirit level: “level on the face of it” will then be good enough. On a classical two-armed balance the effect is again different: on the comparison of two weights there will be no influence, both scales will always hang straight below the measuring point (suspension point). But if the last digits of the difference between the weights are read by a pointer on the arm an error will occur in the sensitivity of this pointing device.

Moving the weighing instrument in height: As the force of attraction (= gravity) between two objects (the earth and the object being weighed) depends on their distance, the measured value will change when the weighing instrument is moved to a different height. By every meter elevation (at moderate heights) the field of gravity changes from sea level -0.3086 N/(mkg). Or as a more practical figure: when the weighing instrument is placed one floor higher (about 4m), a weight of 200g will weigh 0.26mg less.
Moving over a larger distance: The acceleration of gravity g is not the same everywhere on earth: As a consequence of centrifugal forces caused by the earth’s rotation it is less at the equator than at the poles. Also the earth is not a perfect sphere: the distance from the surface to the earth’s center of gravity is larger at the equator than at the poles.
The absolute deviation caused by moving the instrument can be considerable larger than the error of the weights used for adjustment (and larger than the maximum permissible errors for verified weighing instruments). Therefore after moving to a different location readjustment is necessary. Many weighing instruments are fitted with a built-in adjustment weight for this purpose: with this the sensitivity deviation caused by relocating the instrument is easily corrected.

Air currents / drafts: These are often the cause of large random errors, as the effect of air currents depends on unpredictable circumstances like the shape and size of the object being weighed, posture and movements of the operator, or opening doors. Sometimes the metrological demand for the least possible air currents collides with safety requirements: Poisonous substances must be weighed in a fume cupboard. To minimize the effect of ambient air flow, draft shields are used. For balances with a readability of 1mg, an open draft shield, e.g., in the shape of a glass cylinder, will suffice under normal conditions. Balances with a readability of 0.1mg or less require a closed draft shield. Contrary to practical considerations, a draft shield should be as small as possible from a metrological viewpoint, because internal drafts can be generated within a large draft shield chamber.
A special case of air-current effect is what happens when the object being weighed does not have the same temperature as the balance an the surrounding air: Convection currents will make objects that are hotter appear to be lighter and those that are colder, heavier. Only when the shape, surface, and temperature of the object is exactly known, this will be a quantifiable error (see under the chapter “tables” the one about the temperature effects on weights). It is always important to condition an object being weighed to the ambient temperature!

Adsorptive layer of moisture on the surface: Allowing an object sufficient time to reach the same temperature as that of the weighing instrument is also necessary in order for the adsorptive layer of moisture on the surface of the object to equilibrate. Particularly when small objects are weighed on a high-resolution balance, this is essential to obtain reproducible weighing results. To minimize adsorption of humidity, users must avoid soiling or otherwise touching the objects with their hands, for example, as this leaves fingerprints, which can affect the weighing results (a fingerprint does not only have a weight of its own of 0.1 – 1mg, but is also very hygroscopic).

Electrostatic forces: These are generated when nonconductive objects or containers (glass or plastic) are electrically charged, for example, by pouring from one container to another or by air currents. Static electricity can be prevented by maintaining sufficient air humidity (at least 60%), or by using an electrically conductive enclosure that is connected to the weighing pan. Other solutions: Let the sample condition on a metal plate that is connected to the same electrical ground as the balance, ionize the air in and around the balance with an ionizing blower, don’t use a plastic draft shield or sample container: plastics are more easily electrically charged than glass.
Electrostatic forces usually show up as a drift in the weighing results (the balances readout seems to stand still after placing the sample on the pan, but than slowly starts to drift in one direction), or in poor reproducibility in repeated weighing procedures.

Magnetic forces that act on an object being weighed can be counteracted only by carefully demagnetizing the object, by shielding the magnetically active object using soft magnetic materials such as mumetal, or by keeping the object at sufficient distance from all magnetic or magnetizable parts. Weighing instruments with electromagnetic force compensation contain magnetic parts (which are usually well shielded), but even on a mechanical balance the earths magnetic field can have a considerable effect on a magnetized sample. Magnetic forces generally manifest themselves in poor reproducibility of the results and, particularly, in a strong dependency of the position of the object on the weighing pan. An often made mistake is weighing a liquid sample with a magnetic stirrer still in it: This can only be done on a specially prepared scale that always keeps bottle and stirrer in the same position and compensates for the error, or a scale that weighs while stirring; the error than becomes an alternating effect which can be filtered out.

Air buoyancy: An effect which is often underestimated, is air buoyancy: Every object is subject to an upward force, which is equal to the weight  of the displaced air. Therefore, this force will be larger on an object with a large volume (low density), and smaller on an object with high density. A weighing instrument compares two forces; the force exerted on the pan by the object being weighed, and the force exerted by the standard weight that was used to adjust the instrument (or the compensation weights in the case of a mechanical balance). The effect is a difference in display dependent on the density of the object, as depicted in the following graph: