Self-reported information about smoking is imprecise and subject to bias, with accuracy varying according to circumstances. Biochemical assessment gives much clearer indications of the effects of tobacco intake on physiological parameters. As part of a randomized controlled trial, a new point-of-care test for smoking was used as a tool to reduce smoking in pregnancy. Measurements of nicotine metabolites in urine were related to a physiological effect of smoking, notably changes to blood parameters, assessed as a routine part of antenatal care. One hundred and eighty-seven pregnant women attending outpatient antenatal care were initially questioned by a midwife about smoking habits and later questioned by the investigators, during which the test was performed and the results relayed back to the patient. Self-reported smoking habit and cigarette consumption, either reported to midwives or the investigators, were shown to be poor indicators of the effects of smoking on blood parameters. The biochemical assessment of nicotine intake was significantly related to white blood count, haemoglobin concentration, haematocrit, mean cell volume, and mean cell haemoglobin. Red cell count, mean cell haemoglobin concentration, and platelet count were unrelated to nicotine metabolite measurements. We concluded that the new test was a reliable measure of nicotine intake in pregnancy and the results correlated with smoking-related changes to haematological parameters.