Two industrial hygiene studies were conducted at an alkaline battery plant to evaluate worker exposures to manganese dioxide particulate and the effectiveness of filtering facepiece respirators. The work areas studied included the plant's powder-processing tower and press rooms where manganese was blended, compacted with graphite, and inserted into battery cans. Full-shift personal breathing zone monitoring was conducted to estimate manganese dust exposures of press operators, mechanics, and material handlers. In-facepiece and personal breathing zone air sampling pairs were also collected using a program protection factor protocol to estimate the protection provided by the respirators. Particle size evaluations were made using nylon cyclones and Marple personal multi-stage impactors. All samples were analyzed for manganese by inductively coupled argon plasma, atomic emission spectroscopy via NIOSH analytical method 7300 utilizing a modified acid digestion procedure. Fifty-four, full-shift, time-weighted average (TWA) exposures to total manganese ranged from 0.1 to 5.4 milligrams per cubic meter (mg/m3); worker exposures were substantially lower during a follow-up study due to engineering control improvements. Concurrent area sample comparisons of total and respirable manganese revealed that the respirable particulate mass fractions ranged from 6 to 32 percent, and mass median aerodynamic diameters determined from personal breathing zone air samples were mostly greater than 10 micrometers. Fifteen respirator performance evaluations were conducted using Moldex 2200 respirators fitted with 25 millimeter cassettes and light weight sampling probes. Protection factors ranged from 5 to 220, with a geometric mean and standard deviation of 31 and 2.97, respectively. The 5th percentile protection factor estimate was 5, as calculated from the protection factor distribution for this sample set. In 1995, the American Conference of Governmental Industrial Hygienists (ACGIH) lowered the elemental and inorganic manganese dust Threshold Limit Value (TLV) from 5 mg/m3 to 0.2 mg/m3 to address adverse pulmonary and central nervous system effects and male infertility. Although most personal breathing zone concentrations were above 0.2 mg/m3, none of the in-facepiece concentrations exceeded this concentration. Parkinson's-like symptoms have been reported in the literature for high manganese dust and fume exposures, but the importance of low dust exposures for producing neurological effects is uncertain.