and fertiliser. These knapsacks must be filled before placement upon the back of the worker. Spray equipment must be properly maintained. 10. Strategies to reduce pesticide use, such as Integrated Pest Management, should be considered. 11. Increased break time and increased job rotation for particularly labour-intensive jobs associated with physical injury and stress. 12. Workers to be encouraged to take time to work safely and carefully rather than meeting high production quotas. 13. Transport vehicles to have seats, seat belts, roofs and sides offering protection against rollover; separate storage of chemicals for transport to the field; storage place for bush knives during transport; no boarding of moving vehicles allowed. 14. The right to refuse dangerous work. 15. Improved access to impartial health care. This study was supported in part by a grant from the European Uni·on to the South African Farm and Allied Workers Union (SAFAWU). Cherryl Stone and Polly Boniaszczuk assisted in questionnaire and data preparation. Gladness Mnzava was responsible for data entry and Jason Roy conducted most of the statistical analyses. Drs leslie london, Mohamed Jeebhay and ithiseelan Naidoo, and Ms Debbie Budlender of the Community Agency for Social Enquiry, provided insightful comments on earlier versions of this manuscript. Professor Robins was supported on a US Fulbright Scholarship during the period of data collection. Special gratitude is expressed to the leadership, shop stewards and membership of SAFAWU without whose extraordinary efforts this study would not have been possible. References L Segal L. A brutal harvest - the roots of legitimation of violence on farms in South Africa. In: Segal L, ed. The Proiect /(1r tile Study of Violence. Johannesburg: Black Sash. 1990: 1·36. 2. Davies W. We Cry For Our Land. Oxford: Oxfam, 1990; 3. Schoeman B, Plater T. Assaults on farm workers - must they go on and on? CH·n RIghts League Newsletter 1992; 39: ..1,-10. -t. Nordstrom DL, layde PM, Olson KA, Stueland 0, Brand L, Follen Incidence of farm-work-related acute injury in a defined population. Am J fnd Mt'd 1995; 2 : 551- 564. 5. Lee T-Y, Goodwin se, Gibson R\V, Carr WP, Shutske 1. Renier A population- based study of tractor-related injuries: regional rural injury study-I (RRlS-I). J Gccup Ent·jron Hl.'alth 1996; 38: 7 2-793. 6. Van Wendel de Joode Bt\', De Graaf lAM, Wesseling C, Kromhout H. Paraquat exposure of knapsack spray operators on banana plantations in Costa Rica. Int J Occup Envlron Health 1996; 2: 29-1·-30-1. 7. London L Agricultural safety practices on farms in the Western Cape. 5 Afr All.'d / 1994; 84(5): 273-278. 8. Campbell GO. A 'fi\'e-point plan' for improving and maintaining health of farm workers and for increasing \\or output. The South African Sugar Journal 197-4; 58: 437- ++2. 9. loewenson R. Participatory approaches in occupational health research. Paper presented at the 25th International Congress on Occupational Health, Stockholm, Sweden, 1996; 73-79. Accepted J.l Feb 1998. LUNG FUNCTION OF ZIMBABWEA FARM WORKERS EXPOSED TO FLUE CURING AND STACKING OF TOBACCO LEAVES E E Osim, C T Musabayane, J Mufunda -Objectives. To perform lung function tests on tobacco farm workers (TFWs) chronically exposed to flue curing and stacking of tobacco leaves and to compare them with vegetable farm workers (VFWs) who were not exposed to any known air pollutant (control). Detign.Comparativesrudy Setting. Tobacco and vegetable farms. Subjects. 20 TfWs and 30 VFWs. All subjects were male and the mean age, height and weight of the two groups were not significantly different. Outcame measures. Lung function indices. Results. Forced vital capacity (FVq, forced expiratory volume in 1 second (FEV,) and peak expiratory flow rate (pEFR) of the TfWs were 3.28 ± 051 litres, 2.68 ± 0.74 litres and 6.41 ± 2.08litres/second, respectively. These figures were significantly lower than 3.97 ± 0.83 litres, 3.09 ± O.71.litres and 8.62 ± 2.74litres/second, respectively, for the control subjects (P < 0.01,0.05 and 0.01, respectively). mean FEY 1 as a percentage of the FVC (FEVI%) of the TfWs not significantly different from that of the controls. FVC of the TFWs declined with duration of service (r =0.74; P < 0.01). Gmdusion. The results are indicative of restrictive lung defect in the TFWs and may be attributed to long-term exposure to flue curing and stacking of tobacco leaves. The results also suggest the importance of the duration of exposure in the aetiology of lung impairment in this environment. 5 AIr NIl!d f 199; .1127-1131. Oepartllll?llt of Physiology of ZiIllV,'/r,("C, Ham!"e E E Osim, BSc Hons,:-I ,PhD CT Musabayane, B Hons, PhD J Mufunda, \18 ChB, PhD