Introduction. 15 to 25% of women smoke during pregnancy. Scientific evidence suggests that exposure to smoking causes decreased birth weight. The aim of this study was to assess the correlation between smoking during pregnancy, maternal sociodemographic characteristics, and low birth weight. Methods. Data were derived from 1572 questionnaires administered to each woman that gave birth at the Gynecology Teaching Hospital “S. Anna” in Turin (Italy) during the period from 2008 to 2010. Multiple logistic analysis was used to evaluate the association between socio-demographic characteristics and birth weight; the stepwise approach with a “backward elimination” procedure was followed, and the goodness of fit of the model was estimated using the Hosmer-Lemeshow test. Results. The univariate analysis revealed that smoking cigarettes (17%), having a lower educational level (13%), and female sex of the infant (13%) seem to be risk factors, as they increase the risk of having a low birth weight child. Logistic regression analysis showed that gestational age and maternal smoking are the statistically associated variables. Conclusions. The results confirmed that birth weight increases proportionally with the length of the gestational age and that maternal smoking and the child’s sex (female) increase the risk of having a lower birth weight. Logistic regression demonstrated that the association between maternal smoking and low birth weight shows an increased risk for the whole population (OR=2.85), for male (OR=3.45) and for female newborns (OR=2.44) Uvod. Med nosečnostjo kadi 15 % do 25 % žensk. Znanstveni dokazi kažejo, da izpostavljenost kajenju povzroča nižjo porodno težo. Cilj te študije je oceniti povezavo med kajenjem med nosečnostjo, socialnodemografskimi značilnostmi matere in nizko porodno težo. Metode. Podatki so bili pridobljeni iz 1572 vprašalnikov, ki so jih izpolnile vse ženske, ki so v obdobju od leta 2008 do leta 2010 rodile v učni bolnišnici za ginekologijo »S. Anna« v Torinu (Italija). Za oceno povezave med socialno-demografskimi značilnostmi in porodno težo je bila uporabljena multipla logistična analiza; sledil je postopni pristop z »vzvratnim izločanjem« in ocena ustreznosti modela s pomočjo Hosmer- Lemeshowega testa. Rezultati. Univariatna analiza je razkrila, da kajenje cigaret (17 %), nižja stopnja izobrazbe (13 %) in ženski spol dojenčka (13 %) predstavljajo dejavnik tveganja, saj povečujejo tveganje, da bo imel otrok nizko porodno težo. Analiza z logistično regresijo je pokazala, da sta gestacijska starost in kajenje matere statistično povezani spremenljivki. Zaključki. Rezultati potrjujejo, da se porodna teža povečuje sorazmerno z gestacijsko starostjo ter da kajenje matere in spol otroka (ženski) povečujeta tveganje za nižjo porodno težo ob rojstvu. Logistična regresija dokazuje, da povezava med kajenjem matere in nizko porodno težo pomeni povečano tveganje za celotno populacijo (RO = 2,85), tako za novorojenčke moškega (RO = 3,45) kot tudi ženskega spola (RO = 2,44)