OBJECTIVE: To determine changes in standard laboratory measures of dehydration among residents of a nursing home care unit (NHCU) over a 6-month period. DESIGN: A prospective cohort analytic study. SETTING: A 130-bed NHCU in a Department of Veterans Affairs Hospital. PATIENTS: Fifteen infirm but stable male residents (mean age 77 years; range (R) 62-93) on one ward of the NHCU. MAIN OUTCOME MEASURES: We studied prospectively for 6 months the serum osmolality (osm), serum sodium (Na), blood urea nitrogen/creatinine (BUN/Cr) ratios and weight (wt) for 15 patients of the NHCU. None of the patients was acutely ill during the study period or exhibited clinical signs of dehydration. RESULTS: Mean serum osm at baseline: 291.6 mOsm/kg (R 278 to 300); 3 months: 291.5 mOsm/kg (R 276 to 301); 6 months: 291.3 mOsm/kg (R 283-300) were all similar. Forty percent (6/15) of patients had at least one high normal/elevated reading (> or = 295 mOsm/kg) during the study. Three patients (20%) had readings of > or = 300 mOsm/kg, but none of these patients had either concurrent increased serum Na (> or = 146 mmole/L) or BUN/Cr ratios (> or = 25). Mean serum Na at baseline: 143.0 mmole/L (R 139-148); 3 months: 142.1 mmole/L (R 138-149); 6 months: 142.9 mmole/L (R 137-150) were all similar. Sixty percent (9/15) of the patients maintained normal (nl) serum Na levels throughout the study. The relationship between the change in serum Na and serum osm levels from baseline to 6 months was not significant (r = 0.242). BUN/Cr ratios ranged from 12-34 over the study period with 3 of 15 patients (20%) demonstrating elevated ratios consistently throughout the study without clinical evidence of dehydration. Only two patients had both high nl/elevated serum osm and elevated serum Na, although both had nl BUN/Cr ratios. Neither of these patients was thought by staff to be clinically dehydrated. Analysis of variance (ANOVA) indicated none of the laboratory measures changed significantly over time (serum osm: F(2,28) < 1; Na: F(2,28) < 1; BUN/Cr: F(2,28) < 1). There was no significant change in weight between the baseline and six month readings. CONCLUSIONS: These data suggest that in the presence of clinical stability, long-term care residents may have a serum osm in the high normal/elevated range without overt clinical evidence of dehydration, an accompanying elevated Na, or BUN/Cr ratio. This may indicate a different central osm setting for these residents as the serum osm appeared to be stable for each resident over time. These data also suggest that measures of serum osm, Na, and BUN/Cr in the long-term care setting may accurately predict future laboratory values in an individual patient if baseline values are drawn when the patient is not acutely ill.