Angiogenesis, the sprouting of new blood vessels from pre-existing ones, is a key mechanism for solid tumour growth and survival as well as damaged tissue revascularisation. In vitro models of angiogenesis play an important role in elucidating relevant signalling pathways and identifying novel therapeutic strategies. Here we compare two co-culture assays in which fully kinetic measures of vascular tube formation are enabled using live content imaging. Co-cultures of (1) GFP-labelled human umbilical vein endothelial cells (HUVEC) and normal human dermal fibroblasts (NHDF)1 or (2) GFP-labelled endothelial colony forming cells (ECFC) and adipocyte derived stem cells (ADSC)2,3 were created on 96-well assay plates. Vascular tube formation was quantified (tube length, branch points) from fluorescent images (every 6-12h) using an IncuCyte live content imaging system. Exogenous growth factors evoked time and concentration-dependent increases in both assays (e.g. VEGF; HUVEC EC50 28pM, ECFC EC50 54pM). In the HUVEC assay vascular structures were highly branched and formed complex tubular networks. In contrast the ECFC model exhibited more regular cord-like structures; peak cord formation occurred within the first 48h and then stabilised, whereas HUVEC tube structures continued to develop over 2 weeks. From end point antibody staining the tube and cord structures were labelled by vascular markers (CD31, VEGF-R2, n=3). In the ECFC, but not HUVEC model, stromal cells surrounding the vascular structures stained positive for pericyte markers (αSMA and PDGFR-β, n=3). Effects of small molecule pathway inhibitors (e.g. wortmannin, suramin, CCT108159) yielded comparable inhibitory effects across the two models albeit with different temporal profiles. Striking differences in the sensitivity of developing and established tube/cord networks to the anti-VEGF antibody bevacizumab were observed. Neo-angiogenesis was substantially inhibited by bevacizumab (e.g. maximal inhibition at 6.4μg ml-1: HUVEC, 99 ± 0.4%; ECFC, 85 ± 3%) whereas established tube/cord networks were resistant to disruption (HUVEC, 60 ± 3%; ECFC, 30 ± 1%). The γ-secretase inhibitor L-685458 (1-4μM) markedly increased late stage branching in the HUVEC model. Together these data findings further validate these kinetic co-culture models for translational angiogenesis research. The presence of pericyte-like cells in the ECFC/ADSC model may be relevant to the observed resistance of established cords to bevacizumab given that previous studies indicate a role for pericytes in vascular stabilisation. This resistance of established vascular structures to bevacizumab may represent a useful translational paradigm for addressing tumour resistance of anti-VEGF therapies.