Video observations were made of tsetse flies (mainly Glossina pallidipes Austen) as they approached, responded to and left a vertical (1 m square) black target in odour plumes of acetone, or a 4:1:8 mixture of octenol:propylphenol: methylphenol, or acetone plus this “4:1:8” mixture, or in no odour. No differences in mean flight speed or turn size in any of these situations were detected. With the odour source 5 m upwind of the target, the flight tracks of tsetse arriving at and leaving the target were significantly biased towards upwind, highly so when 4:1:8 was present, marginally so in acetone alone. With the source 10 m upwind, the same biases were still present but weaker. Circling flights around the black target were more frequent in acetone plus 4:1:8 than in no odour (26% vs 15%), but in either odour alone were only just significantly more than in no odour. Upwind turning at the target was more frequent (25% vs 17%) in acetone alone that in no odour (though not in 4:1:8 alone). It is concluded that 4:1:8 elicits an upwind anemotactic response comparable in strength to that in CO2, and that acetone elicits a similar response more weakly, but may be more involved in potentiating visual responses.