The potential application for stem cell therapy is vast. Despite a limited understanding of their mechanisms of action, clinical trials assessing stem cells in human stroke are underway. Colony stimulating factors (CSF) such as granulocyte colony stimulating factor (G-CSF), which have been used to mobilise haematopoietic stem cells (HSC), also show promise in treating stroke. Preclinical experiments evaluating the effect of G-CSF in stroke were meta-analysed; G-CSF significantly reduced lesion size in transient but not permanent models of ischaemic stroke. Further studies assessing dose-response, administration time, length of ischaemia and long-term functional recovery are needed. Tracking iron-labelled cells with MRI may help to establish migratory patterns following transplantation. Our systematic review of iron-labelled stem cells in experimental stroke suggests that compounds already licensed for humans (ferumoxide and protamine) may potentially be used in clinical trials. In a phase IIb single-centre randomised controlled trial (n=60), the safety of G-CSF in recent stroke was assessed (STEMS-2). G-CSF appears safe when administered subacutely and may reduce stroke lesion volume. Phase III trials are required to test efficacy. An updated Cochrane review on CSFs in stroke shows that G-CSF was associated with a non-significant reduction in early impairment but had no effect on functional outcome in 6 small studies. In two trials, erythropoietin therapy was significantly associated with death by the end of the trial. It is too early to know whether G-CSF could improve functional outcome in stroke. In 8 recruits randomised into STEMS-2, mobilised CD34+ HSCs were paramagnetically labelled, re-infused and tracked with serial T2* MRI. Post-stroke HSC labelling appears safe and feasible. There is suggestive evidence in one patient that labelled HSCs migrate to the ischaemic lesion. Our in vitro evaluation of CD34+ HSCs has revealed that uptake of superparamagnetic iron oxide (SPIO) is enhanced but not dependent upon a transfection agent. Iron labelling of CD34+ cells in this manner did not affect cell viability or inhibit growth. This methodology could be applied to clinical trials. We have established the expression of G-CSF protein, its receptor (G-CSFR) and CD34 antigen in post-mortem brain tissue from participants recruited to STEMS-2. Areas of angiogenesis and expression of G-CSFR in acute and chronic infarction suggest potential targets for therapy. There are many preclinical studies reporting the effects of stem cells in treating stroke (with a noticeable lack of neutral or negative articles). Despite the wealth of literature there remain many unanswered questions and patients should not undergo stem cell therapy unless it is as part of a well designed clinical trial.