Purpose The authors investigated whether the use of psychotropics and environmental temperature on admission influence hospital length of stay (LOS) and mortality in older medical patients. Methods Clinical and demographic characteristics, Charlson Comorbidity Index, use of psychotropic and nonpsychotropic drugs, hospital LOS, and mortality were retrospectively collected in medical patients 65 years and older (n = 382) admitted to a metropolitan teaching hospital during 5 consecutive heat waves (HWs) between 2007 and 2009. Patients admitted either before or after each HW, matched for HW period, age, and admission day of the week, served as controls (non-HW, n = 1339). Results Total number of psychotropic and nonpsychotropic drugs, Charlson Comorbidity Index, comorbidities, number of daily admissions, LOS, and mortality were similar in the HW and non-HW groups. After adjusting for clinical and demographic confounders, competing risks regression showed that psychotropic use, particularly antipsychotics, predicted increased LOS during non-HW (subdistribution hazard ratio: 95% CI, 0.82, 0.72-0.94; P = 0.003) but not HW (subdistribution hazard ratio: 95% CI, 0.89, 0.69-1.14; P = 0.36) periods. The effect of psychotropics on LOS during normal weather conditions was particularly evident in the old-old subgroup (difference [SE] in coefficients between non-HW and HW periods: -0.52 [0.25], P = 0.036 in patients >80 years; 0.11 [0.19], P = 0.54, in patients 65-80 years). By contrast, psychotropics did not predict hospital mortality during non-HW or HW periods. Conclusions Psychotropic use on admission, particularly antipsychotics, predicted hospital LOS, but not mortality, in older medical patients, particularly those older than 80 years, during normal environmental temperature. However, there was no effect of psychotropics on LOS during extreme heat.