Introduction Indoor microbial toxicity is suspected to cause some building-related symptoms, but supporting epidemiological data are lacking. Objective We examined whether the in vitro toxicity of indoor samples from school buildings was associated with work-related health symptoms (building-related symptoms, BRS). Methods Administrators of the Helsinki City Real Estate Department selected 15 schools for the study, and a questionnaire on symptoms connected to work was sent to the teachers in the selected schools for voluntary completion. The cellular toxicity of classroom samples was determined by testing substances extracted from wiped indoor dust and by testing microbial biomass that was cultured on fallout plates. Boar sperm cells were used as indicator cells, and motility loss was the indicator for toxic effects. The effects were expressed as the half maximal effective concentration (EC50) at which >50% of the exposed boar sperm cells were immobile compared to vehicle control. Results Completed symptom questionnaires were received from 232 teachers [median age, 43 years; 190 (82.3%) women] with a median time of 6 years working at their school. Samples from their classrooms were available and were assessed for cellular toxicity. The Poisson regression model showed that the impact of extracts of surface-wiped school classroom dust on teacher work-related BRS was 2.8-fold (95% CI: 1.6–4.9) higher in classrooms with a toxic threshold EC50 of 6 µg ml−1 versus classrooms with insignificant EC50 values (EC50 >50 µg ml−1); P<0.001. The number of symptoms that were alleviated during vacation was higher in school classrooms with high sperm toxicity compared to less toxic sites; the RR was 1.9 (95% CI: 1.1–3.3, P=0.03) for wiped dust extracts. Conclusions Teachers working in classrooms where the samples showed high sperm toxicity had more BRS. The boar sperm cell motility inhibition assay appears promising as a tool for demonstrating the presence of indoor substances associated with BRS.