Sunitinib is a potent and clinically approved tyrosine kinase inhibitor that can suppress tumour growth by inhibiting angiogenesis. However, conflicting data exist regarding the effects of this drug on the growth of metastases in preclinical models. Here we use 4T1 and RENCA tumour cells, which both form lung metastases in Balb/c mice, to re-address the effects of sunitinib on the progression of metastatic disease in mice. We show that treatment of mice with sunitinib prior to intravenous injection of tumour cells can promote the seeding and growth of 4T1 lung metastases, but not RENCA lung metastases, showing that this effect is cell line dependent. However, increased metastasis occurred only upon administration of a very high sunitinib dose, but not when lower, clinically relevant doses were used. Mechanistically, high dose sunitinib led to a pericyte depletion effect in the lung vasculature that correlated with increased seeding of metastasis. By administering sunitinib to mice after intravenous injection of tumour cells, we demonstrate that while sunitinib does not inhibit the growth of 4T1 lung tumour nodules, it does block the growth of RENCA lung tumour nodules. This contrasting response was correlated with increased myeloid cell recruitment and persistent vascularisation in 4T1 tumours, whereas RENCA tumours recruited less myeloid cells and were more profoundly devascularised upon sunitinib treatment. Finally, we show that progression of 4T1 tumours in sunitinib treated mice results in increased hypoxia and increased glucose metabolism in these tumours and that this is associated with a poor outcome. Taken together, these data suggest that the effects of sunitinib on tumour progression are dose-dependent and tumour model-dependent. These findings have relevance for understanding how anti-angiogenic agents may influence disease progression when used in the adjuvant or metastatic setting in cancer patients.