Gubar (Lymantria dispar) je kao štetnik hrastovih šuma u Hrvatskoj dominirao do 60-tih godina prošloga stoljeća, otkada su njegove populacije u kontinentalnom dijelu u blagom opadanju, sa cikličkim gradacijama svakih 10-11 godina. Zadnje dvije gradacije zabilježene su 2003-2005 i 2013-2014 godine. U radu se opisuje nov način obračuna kritičnog broja jajnih legala gubara, kojim se procjenjuje rizik golobrsta ključan u praktičnom šumarstvu za donošenje odluke o suzbijanju. Nov način obračuna iskazuje kritične brojeve kao broj jajnih legala po hektaru, za razliku od dosadašnjeg koji se temeljio na broju stabala na kojemu se nalazi barem jajno leglo i prikazivao se kao postotak zaraze. Postotak zaraze dijelio je rizik golobrsta u pet klasa, pri čemu je kritična Klasa 5 predstavljala zaraženost više od 50% stabala. Dosadašnjim obračunom nije se uzimala u obzir starost stabla, iako je ona povezana s veličinom krošnje, odnosno količinom lišća. Nadalje nije uziman u obzir broj jaja u jajnom leglu. Nov obračun bazira se na broju gusjenica koje imaju potencijal obrstiti stablo hrasta određene starosti dobiven iz podataka iz literature. Preračunavanjem tih ulaznih podataka korištenjem prosječnog broja jaja u jednom jajnom leglu gubara, koji je dobiven iz 50 uzoraka jajnih legala, te normale koja definira broj stabla hrasta lužnjaka po hektaru, dobivena je formula koji izračunava kritični broj jajnih legala gubara po hektaru. Na osnovi tih rezultata formirane su tri 3 kategorije rizika temeljene na broju gusjenica koji obrste stablo određene starosti. Svaka od njih različito definira potrebu za suzbijanje gubara: Kategorija I- treba tretirati, Kategorija II- iznimno tretirati uz specifičan razlog i Kategorija III ne treba tretirati. Kritična Kategorija I uspoređena je s Klasom 5, odnosno Kategorija II s Klasom IV na podacima iz gradacije 2004. i 2013. godine. Rezultati ukazuju kako je primjenom novog obračuna trošak suzbijanja smanjen za 25%. Uz evidentne prednosti ipak su i u novom obračunu uočene slabosti vezane uz nepoznavanje nekoliko važnih varijabli: stvarni broj stabala po hektaru i stvarni broj intaktnih gusjenica koji se dobiva oduzimanjem parazitiranih i sterilnih jaja od prosječnog broja jaja u leglu. Još jedna slabo poznata varijabla prosječna količina lišća koju jedna gusjenica može obrstiti te količinu lišća u krošnji hrastova određene starosti koju treba u budućim istraživanjima kvalitetnije definirati. Gypsy moth, Lymantria dispar was the dominat oak forest pest in Croatia untill 1960s. After this period population particularly in the continental part decreased with cyclical outbreaks every 10-11 years. Last two outbreaks occurred in 2003-2005, and in 2013-2014. In this study a new calculation of critical number of gypsy moth egg masses has been presented in order to estimate the risk of defoliation which is crucial for the decision about application of necessary control measures. The calculation shows critical numbers as number of egg mases per ha which differ from the previous calculations based on number of trees which carry at least 1 egg mass shown in percentages. Hitherto 5 Classes were known out of which the 5th was critical. This Class presented 50% of trees which carry at least 1 egg mas and it was assessed as critical. In such calculation the age of the trees was not taken in account although the amount of leaves in the canopy strongly depend on that. Furthermore the number of eggs in the egg masses was also ignored. New calculation is based on numbers of caterpillars which have the potential to defoliate an oak tree of different ages, obtained from literature. The average number of eggs in one egg mass was calculated out of 50 sampled and analysed egg masses. This data together with the normal distribution of oak trees per ha build the basis for the formula. Based on that 3 Categories were formulated which show the potential for defoliation and help to make the decision for applying control measures: Category I – control measures needed; Category II – control measures needed only if there is a special reason; Category III – no control measures needed. Critical numbers of old and new calculation have been compared in years of outbreaks. Results between new and old calculation show 25% difference which directly means 25% less area that needs treatment with insecticides. Although the new method has advantages there is still some limitation in the calculation considering that important variables are ignored. The real number of trees per ha and the actual number of intact caterpillars should be used for more accurate calculation and further reduction of forest area that needs protection. Future research should concentrate on precise calculations of leaf weight consumed by average caterpillar in the relationship to crown volume both for oak and common beech. These values in combination with the number of leaves in the particular forest would make the decisions on use of insecticides more accurate which would bring additional substantial financial savings.