OBJECTIVE The function of jet nebulizers has been measured traditionally by gravimetric methods, i.e., by weighing nebulizers before and after nebulization. Newer techniques measure aerosol output directly by analyzing aerosolized drug or tracer, i.e., radioactive 99mTc. Because of evaporation, the equivalence of these methods is uncertain. The aim of this study was to determine if the gravimetric method is an accurate measure of aerosol production under different conditions of aerosol generation (i.e., nebulizer type, flow rate, pressure, volume fill, and concentration of solution used to nebulize a drug). METHODS In the first phase of the study, we measured the aerosol output of nine commercially available jet nebulizers (AvaNeb; Up-Draft-Hudson RCI; Cirrus-Intersurgical Inc; DeVilbiss 646-DeVilbiss; Powermist-Hospitak, Inc; Respirgard II-Marquest Medical Products; Seamless-Seamless/Dart Respiratory; Salter; Salter Labs; Airlife-Baxter Health Care) run under commonly used conditions (2.5 mL volume fill, 2.0 mL normal saline solvent, 0.5 mL albuterol, flow of 6 L/min, and pressures averaging 15.0 +/- 2.3 [mean +/- SD] pounds per square inch [on the] gauge [psig] provided by a DeVilbiss PulmoAide compressor) with simultaneously measured gravimetrics and filtered radioactivity. Each nebulizer was run to dryness with data acquired every 2 min. The change in the weight of the nebulizer and radioactivity captured on the filter were expressed as percentages of the total in the nebulizer solution. In the second phase of the study, the experiments were repeated using the same nebulizers with a volume fill of 5 mL (diluted to half normal saline solution plus albuterol), flow of 10 L/min, and pressures of 35.6 +/- 8.8 psig. RESULTS The cumulative (sum of all 2-min runs) weight loss for each individual nebulizer ranged from 25.00 to 64.55% and cumulative aerosol captured varied from 12.63 to 38.76%. While different, the weight loss and aerosol captured were closely correlated (y = -0.62 + 0.62x; r = 0.961, p < 0.0001). Changing volume fill and concentration of solvent did not affect this correlation (p = 0.921 and 0.373, respectively). However, changing flow from 6 L/min to 10 L/min significantly (p = 0.02) affected the relationship (y = -3.80 + 0.83x; r = 0.969, p < 0.001). CONCLUSIONS When compared with direct methods such as filtering generated particles, the gravimetric method of assessing nebulizer function overestimates aerosol output by 1.8 +/- 0.18 times, presumably because of the loss of solvent during nebulization. However, the relationship between methods is predictable and appears unaffected by changing the type of nebulizer, volume fill, and concentration of solvent. Changes in nebulizer flow and pressure significantly affected the correlation. Gravimetric methods can be used as simple and convenient screening techniques for comparing jet nebulizers under a wide range of experimental conditions.