The purpose of this study were to analyze work function and labor productivity in school foodservice systems through work sampling methodology. Conventional foodservice were classified into 5 group depending on the number of meals served. Commissary school foodservices were also classified into 5 group by cluster analysis using the number of meals served, the number of satellite schools, and the duration of time for delivery of food. Work measurement through work sampling methodology was conducted in at 5 conventional and 21 commissary food -services during 3 consecutive days from September to October in 1995. Results from work measurement through work sampling methodology were as followed : The most prevalent work functions was cleaning (26.5%) and then processing (25.1%) in conventional while it was processing (30.9%) and then cleaning(25.2%) in commissary school foodservice. Delay was 22.9% and 19.7% respectively. Mean labor minutes per meal of conventional and commissary foodservics were 4.57 and 4.09 minutes, respectively : no significant difference in labor minutes per meal existed between the two systems. but mean labor minutes per meal of commissary foodsevices(1.79min) was significantly lower than that of conventional foodservice(2.33min) during work time before service (p<0.05). Productivity was significantly lower in foodservices which served less than 400 servings ; n significant differences existed among 401-70, 701-1,100 and 1,101-1,500 meals. The highest productivity was in conventional school foodservices which served 1,501-1,900meals. Labor minutes per meal of commissary school foodservice which served less than 400meals per day was significantly lower than those of foodservice which served 401-1,900meals (p<0.05). Labor minutes per meal, preparation and cleaning were positively correlation in two school foodservice systems.