Urdangarin was accused of using his royal connections to generate business income used for private spending.
The case began in 2010 and became symbolic of perceived corruption among Spain`s elites, including the royals.
There were 16 other defendants in the case, including former government minister Jaume Matas, who was sentenced to three years and eight months.
Urdangarin`s former business partner, Diego Torres, was given eight years and six months. Nine defendants in all were acquitted.
Princess Cristina, who now lives in Switzerland, was the first member of Spain`s royal family to go on trial since the monarchy`s restoration in 1975.
In 2015, King Felipe stripped her and Urdangarin, 49, of their titles as Duke and Duchess of Palma de Mallorca.
Tax havens
Although she was absolved, Princess Cristina will still have to pay a fine of €265,000 (£227,000; $282,000) as she has civil responsibility for benefiting, albeit unknowingly, from illegal gains.
The couple were not in court for the verdict, which is subject to appeal. Both had denied wrongdoing.
The princess had been accused of being an accessory to tax fraud.
Public prosecutors had declined to press charges against her but the three judges agreed to continue with the prosecution using evidence filed by the anti-corruption group Manos Limpias, meaning "Clean Hands".
She told the court last March that her husband was in charge of family finances, saying: "I didn`t get involved in that."
Urdangarin, a former Olympic handball medallist, had faced a slew of more serious charges.
He was accused of using the non-profit Noos Institute sports foundation he ran as a vehicle to win falsely inflated contracts from regional government bodies, before channelling the money to personal accounts via tax havens.
Noos is alleged to have received more than €6m (£4.4m; $6.5m) of public money, most of it from the Balearic Islands and Valencia regional governments.
One of the companies said to have received money, real estate firm Aizoon, was jointly owned by Cristina and Urdangarin.
More about:
